Ep 28 — Environmental Monitoring: Atmospheric Gas Monitoring and Emission Source Identification

Series: Encyclopedia of Infrared Spectroscopy: From Principles to Practice
Chapter: Part III · Intermediate — Industry Applications
Target Audience: Environmental monitoring technicians, atmospheric chemistry researchers, emission monitoring engineers
Prerequisites: Ep 13 (Transmission Method and Gas Cells), Ep 20 (Quantitative Analysis), Ep 27 (Environmental Monitoring: Water Quality and Soil)
Reading Time: Approximately 48 minutes

Introduction: The "Fingerprint" of a Chimney

In an industrial area in 2024, residents complained of a pungent odor at night. When environmental enforcement officers arrived, several chimneys were emitting normally, and it was impossible to determine which factory was violating regulations [1].

The monitoring team deployed an Open-Path FTIR system, directing the infrared beam through the plume and receiving it 500 meters away. Within just 5 minutes of measurement, the spectrum showed characteristic absorptions of toluene, dichloromethane, and ethyl acetate—the "fingerprint" emissions of the cleaning workshop of an electronics factory [1].

"Open-Path FTIR can identify and quantify dozens of gases simultaneously, in real-time, without sampling."
— EPA OTM 10 Open-Path FTIR Method Guide [2]

Gas FTIR is the "super nose" of atmospheric monitoring—it can [1][2][3]:

• Identify dozens of gases simultaneously (VOCs, CO, NOₓ, SO₂, NH₃, etc.)
• Quantify in real time (second-level response)
• Perform sampling-free analysis (direct measurement via open path)
• Monitor over long distances (path length up to hundreds of meters to kilometers)

This episode delves into the principles, technical systems, and emission source identification practices of gas FTIR, focusing on the core technical challenge of subtracting water vapor and CO₂ interference.

1. Physical Basis of Gas FTIR

1.1 Infrared Activity of Gas Molecules

Infrared absorption in the gas phase differs from the condensed phase [3][4]:

① Very Narrow Spectral Lines (Line Spectrum):

• Gas molecules rotate freely, producing a vibrational-rotational spectrum.
• Each vibrational band consists of dense rotational lines.
• Line width is approximately 0.1–0.5 cm⁻¹ (affected by pressure broadening and Doppler broadening).

② Selection Rules:

• Vibrational transitions: Δv = ±1
• Rotational transitions: ΔJ = ±1 (P branch), ΔJ = +1 (R branch), ΔJ = 0 (Q branch)
• Form characteristic PQR three-branch structure.

③ Concentration–Path Length Relationship (Beer-Lambert):
$$A = \varepsilon \cdot c \cdot L$$

• Gas concentration c is usually expressed in ppm or mg/m³.
• Path length L is a key variable—long path length is the core strategy for gas detection.

🔗 Extension: For peak positions of common gaseous pollutants, in addition to consulting HITRAN, you can also cross-reference functional groups/peaks at ftir.fun: water, adsorbed/CO-related, nitro/nitrogen-oxygen related, peak query entry. Note that the fine vibrational-rotational structure in the gas phase is not exactly equivalent to the condensed-phase functional group page.

🔗 Extension: Water vapor (H₂O), as a major atmospheric component, has an extremely complex vibrational-rotational spectrum and is the main interference in gas FTIR analysis. For detailed analysis, see ftir.fun water molecule functional group page.

1.2 Relationship Between Detection Limit and Path Length

The detection limit of gas FTIR depends on path length [3][4]:

Table 1: Relationship between path length and detection limit (Data source: Griffiths & de Haseth [3])

2. Long-Path Gas Cells

To achieve the effective path length required for trace gas detection, gas cells increase path length through multiple reflections [3][5].

2.1 White Cell

The White cell, invented by John White in 1942, is the most classic long-path gas cell [3][5]:

Structure: Three concave mirrors (spherical mirrors)

• Incident light is focused by the objective mirror and reflects repeatedly between two "corner mirrors."
• Finally exits through the exit slit.

Path Length Calculation:
$$L = N \cdot d$$

• N: number of reflections (typically 4–52, up to 100+)
• d: mirror spacing (typically 0.5–3 m)
• Typical path length: 2–100 m (up to ~300 m)

  White cell light path (4 reflections schematic):

      Incident →  ┌───┐
                   │ Mirror A │ ─→ Mirror B
                   └───┘     ↓
                        ←────┘
                        ↓
                     Mirror C ─→ Exit

Features [3][5]:

• Relatively large volume (tens of liters)
• Adjustable path length (by changing number of reflections)
• Sensitive to optical alignment
• Suitable for high-precision laboratory measurements

2.2 Herriott Cell

The Herriott cell, proposed by Herriott & Schulte in 1964, is a compact improvement over the White cell [3][5]:

Structure: Two concave mirrors (spherical or cylindrical)

• Light enters through an off-axis hole in one mirror.
• Reflects multiple times between the two mirrors, forming an elliptical spot pattern.
• Finally exits through an exit hole.

Path Length Calculation:
$$L = N \cdot 2d$$

• N: number of reflections (typically 20–100+)
• d: mirror spacing (typically 10–30 cm)
• Typical path length: 5–50 m (up to ~200 m)

Features [3][5]:

• Small volume (0.5–5 L) — compact
• Fixed path length (determined at design)
• Good optical stability
• Suitable for portable, online monitoring

📷 Figure 1: Comparison of White cell and Herriott cell light paths
Source: Thorlabs Long-Path Gas Cell Technical Data [5]
https://www.thorlabs.com/newg…

2.3 Comparison of Mainstream Long-Path Gas Cells

Table 2: Comparison of White cell vs. Herriott cell

2.4 Key Gas Cell Operating Points

① Cell Materials [3]:

• Windows: KBr (mid-IR), ZnSe (moisture resistant), CaF₂ (high pressure)
• Cell body: Stainless steel or glass (nickel-plated for corrosion resistance)

② Temperature Control:

• Gas concentration is affected by temperature (ideal gas law PV=nRT)
• High-precision measurements require temperature control to ±0.1 °C

③ Pressure and Purge:

• Standard measurement: atmospheric pressure (101.3 kPa)
• Reducing pressure reduces spectral line broadening (improves resolution)
• Purge with N₂ before measurement to remove background gases

3. Infrared Characteristics of Atmospheric Pollutants

3.1 Infrared Absorption of Major Atmospheric Pollutants

| Pollutant | Molecular Formula | Main Absorption Bands (cm⁻¹) | Typical Concentration | Source |

|--------|--------|------------------|---------|------|
| Carbon monoxide | CO | 2143 (fundamental) | 0.1–10 ppm | Combustion, motor vehicles |
| Nitrogen dioxide | NO₂ | 1620, 2900 | 0.01–1 ppm | Combustion, motor vehicles |
| Nitric oxide | NO | 1876 | 0.01–1 ppm | Combustion |
| Sulfur dioxide | SO₂ | 1361, 1151 | 0.01–1 ppm | Coal burning, smelting |
| Ammonia | NH₃ | 950, 3334 | 0.01–0.5 ppm | Agriculture, chemical industry |
| Methane | CH₄ | 3017, 2917 | 1.8 ppm (background) | Natural gas, agriculture |
| N₂O | N₂O | 1285, 2224 | 0.3 ppm (background) | Agriculture, industry |
| Benzene | C₆H₆ | 675, 1038, 1485, 1815 | 1–100 ppb | Chemical industry, solvents |
| Toluene | C₇H₈ | 728, 1030 | 1–100 ppb | Coatings, solvents |
| Formaldehyde | HCHO | 2780, 1746 | 1–50 ppb | Indoor decoration, combustion |
| Methanol | CH₃OH | 1033, 1345 | 1–100 ppb | Chemical industry, solvents |
| Hydrogen chloride | HCl | 2886 | 0.1–10 ppm | Incineration, chemical industry |

Table 3: Infrared characteristic absorptions of major air pollutants (data sources: EPA FTIR spectral library [2]; HITRAN database [4])

3.2 Fingerprint Region and Quantitative Region

The infrared spectra of gas molecules can be divided into two types of regions [2][3]:

① Quantitative Region:

• Select "windows" where the target gas has strong absorption and other gases have weak absorption
• Example: 2143 cm⁻¹ region for CO (weak interference from H₂O, CO₂)
• Used for quantitative analysis

② Fingerprint Region:

• Dense vibration-rotation lines of multiple gases
• Used for qualitative identification ("recognize at sight")
• Difficult for quantification (severe overlap)

💡 Practical tip: Gas quantitative analysis commonly uses multivariate chemometric methods (CLS, PLS), which can simultaneously resolve multiple gases [2][3]. EPA OTM 10 method uses Classical Least Squares (CLS).

4. Open-Path FTIR

4.1 Principle

Open-Path FTIR (OP-FTIR) does not use a gas cell; instead, the infrared beam is directed directly through the open atmosphere [1][2][6]:

  OP-FTIR configuration (active):

  IR source ─→ Telescope ──→ [Atmospheric path 100–500 m] ──→ Telescope ─→ Detector
                                 ↑
                           Plume/pollution cloud crossing the path

Two configurations [1][2][6]:

• Active: Built-in IR source, receiver at far end—fixed path length, strong signal
• Passive: Uses natural thermal radiation (sunlight, terrestrial radiation)—no source needed, but lower sensitivity

4.2 Measurement Modes

① Point monitoring [1]:

• Source and detector at the same end (retroreflector returns from far end)
• Measures average concentration over a fixed path
• Suitable for fence line, workshop boundary monitoring

② Tomographic scanning [6]:

• Multiple intersecting paths
• Reconstructs 2D concentration distribution maps
• Suitable for overall monitoring of industrial parks

③ Plume crossing [1][2]:

• Beam passes through the plume emitted from a stack
• Directly measures emission concentration
• No sampling required, avoids sample distortion

4.3 Advantages and Limitations of OP-FTIR

Advantages [1][2][6]:

• No sampling: Avoids adsorption and reaction losses during sampling
• Real-time: Second-level response
• Multi-component simultaneous: One measurement covers dozens of gases
• Area-averaged: Measures average concentration over the path, good representativeness
• Long distance: Path length up to hundreds of meters to kilometers

Limitations [1][6]:

• Detection limit: 1–2 orders of magnitude higher than gas cells (affected by atmospheric turbulence)
• Weather dependent: Rain, fog, snow severely affect the path
• Complex quantification: Requires precise knowledge of path length and temperature/pressure distribution
• High equipment cost: Telescopes, tracking systems are expensive

📷 Figure 2: OP-FTIR field monitoring of stack emissions
Source: EPA OTM 10 application case [2]
https://www3.epa.gov/ttnemc01…

5. Continuous Emission Monitoring Systems (CEMS) for Stack Gas

5.1 FTIR-CEMS System Components

FTIR-CEMS (Continuous Emission Monitoring System) is a standard system applying FTIR to continuous monitoring of fixed source flue gas [2][7]:

  FTIR-CEMS process:

  Stack ─→ Sampling probe (heated line 180°C)
              ↓
          Particulate filter
              ↓
          Heated transfer line (to avoid condensation)
              ↓
          FTIR gas cell (heated to 180°C)
              ↓
          Detector + data processing
              ↓
          Multi-gas quantitative results → Environmental agency data platform

① Sampling system [2][7]:

• Sampling probe: Stainless steel, front-end filter (> 2 μm particulate)
• Heated line: 180 °C constant temperature, prevents water vapor condensation (avoids SO₂, HCl dissolution loss)
• Calibration gas port: Periodic introduction of standard gases for calibration

② FTIR analysis unit [2][7]:

• Gas cell: 10–20 m path length, heated to 180 °C
• Resolution: 1–2 cm⁻¹ (line spectra of gases require high resolution)
• Scan frequency: One spectrum every 1–5 minutes

③ Data processing [2][7]:

• CLS (Classical Least Squares) multivariate quantification
• Built-in NIST / EPA gas reference spectral libraries
• Automatic subtraction of water vapor and CO₂ background

5.2 CEMS Measurement Parameters

US EPA Method 320 specifies the main pollutants measurable by FTIR-CEMS [2]:

Table 4: Typical measurement parameters of FTIR-CEMS (data source: EPA Method 320 [2])

5.3 Chinese HJ 75 Standard

China's HJ 75-2017 "Technical Specification for Continuous Emission Monitoring of Stationary Source Flue Gas (SO₂, NOₓ, Particulate Matter)" allows the use of extractive FTIR or direct penetration FTIR [7]:

Extractive [7]:

• Heated sampling → FTIR gas cell measurement
• Advantages: Compatible with existing CEMS
• Limitations: Sampling losses (particulate adsorption, condensation)

In-situ (direct penetration) [7]:

• Source and detector installed on opposite sides of the stack
• Beam passes directly through stack gas
• Advantages: No sampling losses
• Limitations: Alignment difficulties, complex maintenance

6. Water Vapor and CO₂ Interference Subtraction

6.1 Atmospheric Background Interference Issue

Water vapor (0.1–4%) and CO₂ (0.04%) in the atmosphere are the main interferences in FTIR gas analysis [3][4][8]:

Water vapor interference regions [3][4]:

• 3400–4000 cm⁻¹ (O-H stretching) — overwhelms NH₃, HCHO
• 1300–2000 cm⁻¹ (H-O-H bending and combination) — overwhelms NO₂, SO₂, HCHO
• 5000–5600 cm⁻¹ (combination) — near-infrared region interference

CO₂ interference regions [3][4]:

• 667–730 cm⁻¹ (bending) — overwhelms VOCs
• 2280–2390 cm⁻¹ (asymmetric stretching) — overwhelms CO

🔗 Extension: For the complex rovibrational spectral analysis of water vapor, see ftir.fun water functional group page. Understanding water absorption bands is fundamental for interference subtraction.

6.2 Conventional Subtraction Methods

① Subtraction method [3][8]:

• Measure background of clean air (or N₂)
• Subtract water vapor and CO₂ from sample spectrum
• Limitation: Water vapor concentration varies with time, background hard to match precisely

② Reference spectrum fitting [3][8]:

• Compute high-resolution reference spectra of water vapor and CO₂ from HITRAN database
• Fit to sample spectrum, subtract contributions
• Limitation: Requires accurate knowledge of temperature, pressure, path length

③ Window selection [2]:

• Select "window" regions with weak water vapor absorption for quantification
• Example: CO quantification at 2143 cm⁻¹ (weak water absorption)
• Limitation: Limits the range of measurable gases

6.3 VaporFit: Automatic Atmospheric Subtraction Tool

VaporFit is an open-source atmospheric subtraction software released in 2025 in Phys. Chem. Chem. Phys. [8]:

"VaporFit enables automatic subtraction of atmospheric water vapor and CO₂ from FTIR spectra, significantly improving the detection of trace gases."
—— Bruzdowski et al., PCCP 2025 [8]

Core Algorithm [8]:

1. Modeling: Generate reference spectra of water vapor and CO₂ based on HITRAN database
2. Fitting: Fit reference spectra to sample spectrum using nonlinear least squares
3. Subtraction: Subtract fitted water vapor and CO₂ contributions from sample spectrum
4. Iteration: Iteratively optimize fit parameters (concentration, temperature, instrumental line shape)

Features [8]:

• Automatic: No manual parameter adjustment needed
• Open-source: Python implementation, free to use
• Universal: Supports any FTIR gas spectrum
• High precision: Subtraction residual < 1%

GitHub Repository: https://github.com/piobruzd/V… [8]

Usage Example [8]:

from vaporfit import VaporFit

# Load sample spectrum
vf = VaporFit(spectrum_file="sample.spc")

# Automatically fit and subtract water vapor and CO₂
vf.fit_atmosphere()
corrected = vf.get_corrected_spectrum()

# Output subtracted spectrum
corrected.save("corrected.spc")

6.4 Comparison of Subtraction Effects

A comparison of an air sample containing 0.5 ppm HCHO before and after subtraction [8]:

Table 5: VaporFit atmospheric subtraction effect (data source: Bruzdowski et al. [8])

💡 Key Insight: Atmospheric subtraction is a fundamental skill in gas FTIR analysis. Automated tools like VaporFit have democratized this process, which previously required expert experience, enabling non-spectroscopists to obtain high-quality trace gas quantification results [8].

📷 Figure 3: Comparison of spectra before and after water vapor subtraction using VaporFit
Source: Bruzdowski et al., PCCP 2025 [8]
https://github.com/piobruzd/V…

7. Case Study: Identification of VOCs Emission Sources in an Industrial Park

7.1 Background

A chemical industrial park covering 5 km² contains over 30 enterprises (coatings, plastics, pharmaceutical intermediates, etc.). Residents have repeatedly complained about odors, but due to the large number of enterprises, it is difficult to locate the main emission sources [1][6].

7.2 OP-FTIR Monitoring Plan

Researchers deployed 4 OP-FTIR monitoring points at the park boundary and interior [1][6]:

① North point: Path length 300 m, monitoring north boundary
② East point: Path length 200 m, monitoring boundary of a coating factory
③ South point: Path length 500 m, monitoring south boundary (residential area side)
④ Center point: Path length 400 m, monitoring park center

7.3 Monitoring Results

24-hour continuous monitoring identified 15 VOCs, with main findings [1]:

Table 6: VOCs emission characteristics identified by OP-FTIR (data source: monitoring report [1])

7.4 Source Identification Value

The chemical fingerprints from OP-FTIR enabled precise source identification [1][6]:

• Toluene + Ethyl acetate combination → Coating factory characteristic (solvent formulation)
• Methanol + Acetone combination → Byproducts of pharmaceutical intermediate synthesis
• Dichloromethane single → Cleaning agent for plastic factory
• Styrene single → Unreacted monomer from PS polymerization plant

💡 Key Insight: Each factory has its unique "VOCs fingerprint" — FTIR's ability to simultaneously identify multiple components makes this fingerprint-based source identification possible, which cannot be achieved by single-target sensors (e.g., NDIR for CO) [1].

8. Practical Experience in Gas FTIR

8.1 Resolution Selection

Resolution selection in gas FTIR is critical [2][3]:

Table 7: Resolution selection guide

⚠️ Note: The rovibrational linewidth of gas molecules is about 0.1–0.5 cm⁻¹; resolution worse than 1 cm⁻¹ significantly reduces quantitative accuracy. However, higher resolution increases scan time and decreases SNR. Trade-offs must be made based on the application [2][3].

8.2 Instrument Line Shape (ILS)

Instrument Line Shape (ILS) is a function describing the resolving power of an FTIR instrument [3]:

• Determined by resolution (maximum optical path difference of the moving mirror)
• Also affected by apodization function
• ILS determines the degree of spectral line broadening

ILS Correction [3]:

• Measure known narrow lines (e.g., HBr, HCl gas)
• Fit instrument line shape parameters
• Correct ILS effects in quantitative analysis

8.3 Calibration Gases and Spectral Libraries

① Standard Gas Calibration [2]:

• Purchase NIST-traceable standard gases
• Measure instrument response at known concentrations
• Establish working curves

② Reference Spectral Libraries [2][4]:

• NIST WebBook: Gas-phase IR spectra
• EPA/EMC FTIR Reference Spectra: For HAPs etc., publicly available with scales typically on the order of hundreds
• PNNL Library: Gas-phase FTIR spectra of 480 organic compounds
• HITRAN: High-resolution gas-phase spectra of 47 small molecules

🔗 Extension: Ep 55 will summarize free spectral databases in detail.

Summary of This Episode

Review Questions

1. A gas FTIR spectrum shows prominent rotation-vibration lines near 2143 cm⁻¹. Infer the possible gas and explain why this region is suitable for quantification.
2. What are the advantages and disadvantages of the White cell and Herriott cell? Which would you choose for designing a portable atmospheric monitoring scheme? Why?
3. OP-FTIR signal drops significantly in rainy or foggy weather. Explain the reason from an optical perspective and propose countermeasures.
4. In a flue gas FTIR-CEMS measurement of SO₂, the result is found to be 30% higher. What are the possible causes? How to troubleshoot?
5. After subtracting water vapor with VaporFit, a weak peak appears at 1746 cm⁻¹ in a sample. What pollutant might it be? How to confirm?
6. Why does gas FTIR quantification require higher resolution (1–2 cm⁻¹) than liquid/solid? Explain from the perspective of molecular spectroscopy.
7. Design an OP-FTIR monitoring scheme to locate nighttime fugitive emissions from a chemical plant, including steps for site selection, measurement, and source identification.

References

Standards and Methods

[1] US EPA. "OTM 10: Open-Path Fourier Transform Infrared Spectroscopy (OP-FTIR) for Source Emission Monitoring." EPA Emission Measurement Center.
https://www3.epa.gov/ttnemc01…

[2] US EPA. "Method 320: Measurement of Vapor Phase Organic and Inorganic Emissions by Extractive Fourier Transform Infrared (FTIR) Spectroscopy." EPA CFR 40 Part 63, Appendix A.
https://www3.epa.gov/ttn/emc/…

[7] Ministry of Ecology and Environment of the People's Republic of China. HJ 75-2017 Technical Specification for Continuous Emission Monitoring of Stationary Source Flue Gas (SO₂, NOₓ, Particulate Matter). China Environmental Science Press, 2017.

Principles and Reviews

[3] Griffiths PR, de Haseth JA. Fourier Transform Infrared Spectrometry. 2nd ed. Wiley, 2007. Chapter 10: "Gas-Phase Spectroscopy"; Chapter 19: "Atmospheric Monitoring."

[4] Rothman LS et al. "The HITRAN 2012 Molecular Spectroscopic Database." Journal of Quantitative Spectroscopy and Radiative Transfer, 2013, 130: 4–50. DOI:10.1016/j.jqsrt.2013.07.002.
https://hitran.org/

[5] Thorlabs. "Multipass Gas Cells for FTIR Spectroscopy." Technical Documentation.
https://www.thorlabs.com/newg…

[6] Russwurm GM, Childers JW. "Open-Path Fourier Transform Infrared Spectroscopy." Handbook of Vibrational Spectroscopy, Wiley, 2006. DOI:10.1002/0470027320.s4604.

Atmospheric Subtraction Tools

[8] Bruzdowski P et al. "VaporFit: An Open-Source Tool for Automatic Subtraction of Atmospheric Water Vapor and CO₂ from FTIR Spectra." Physical Chemistry Chemical Physics, 2025, 27. DOI:10.1039/D5CP01007A.
GitHub: https://github.com/piobruzd/V…

Database Resources

[ftir.fun] ftir.fun Infrared Spectral Database. Water molecule functional group page:
https://ftir.fun/ir/group/wat…

[NIST] NIST Chemistry WebBook. Gas-Phase IR Spectra.
https://webbook.nist.gov/chem…

[PNNL] Pacific Northwest National Laboratory. "FTIR Spectra of Vapor Phase Chemicals."
https://secure2.pnl.gov/nsd/n…

Next Episode Preview: Ep 29 — Petrochemical: Oil Analysis and Fuel Quality
We will turn to the petrochemical industry, explaining the infrared characteristics of gasoline, diesel, and lubricating oils, quantification methods for fuel oxygenates (methanol, ethanol, MTBE), lubricant aging monitoring (oxidation, nitration, sulfation values), and industry standards ASTM E2412 and JOAP.

This article is licensed under CC BY-NC-SA 4.0. Images are from public domain or online resources with credited sources, copyrights belong to original authors.